{"id":"https://openalex.org/W7160887789","doi":"https://doi.org/10.48550/arxiv.2605.10094","title":"Retrieve-then-Steer: Online Success Memory for Test-Time Adaptation of Generative VLAs","display_name":"Retrieve-then-Steer: Online Success Memory for Test-Time Adaptation of Generative VLAs","publication_year":2026,"publication_date":"2026-05-11","ids":{"openalex":"https://openalex.org/W7160887789","doi":"https://doi.org/10.48550/arxiv.2605.10094"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.10094","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10094","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2605.10094","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101196282","display_name":"Jianchao Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Jianchao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114178241","display_name":"Huoren Yang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yang, Huoren","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135954072","display_name":"Hu Yusong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Yusong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135927052","display_name":"Yuyang Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Yuyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135932237","display_name":"Qiguan Ou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ou, Qiguan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135943822","display_name":"Cong Wan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wan, Cong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135972421","display_name":"SongLin Dong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dong, SongLin","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136000891","display_name":"Zhiheng Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Zhiheng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135974016","display_name":"Yihong Gong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gong, Yihong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":9,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.39399999380111694,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.39399999380111694,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10653","display_name":"Robot Manipulation and Learning","score":0.3822000026702881,"subfield":{"id":"https://openalex.org/subfields/2207","display_name":"Control and Systems Engineering"},"field":{"id":"https://openalex.org/fields/22","display_name":"Engineering"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.03999999910593033,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/task","display_name":"Task (project management)","score":0.6590999960899353},{"id":"https://openalex.org/keywords/reliability","display_name":"Reliability (semiconductor)","score":0.6373999714851379},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.6151000261306763},{"id":"https://openalex.org/keywords/exploit","display_name":"Exploit","score":0.6140999794006348},{"id":"https://openalex.org/keywords/reuse","display_name":"Reuse","score":0.5950000286102295},{"id":"https://openalex.org/keywords/action","display_name":"Action (physics)","score":0.5921000242233276},{"id":"https://openalex.org/keywords/generative-grammar","display_name":"Generative grammar","score":0.5527999997138977},{"id":"https://openalex.org/keywords/robot","display_name":"Robot","score":0.4611000120639801}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.723800003528595},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.6590999960899353},{"id":"https://openalex.org/C43214815","wikidata":"https://www.wikidata.org/wiki/Q7310987","display_name":"Reliability (semiconductor)","level":3,"score":0.6373999714851379},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6180999875068665},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.6151000261306763},{"id":"https://openalex.org/C165696696","wikidata":"https://www.wikidata.org/wiki/Q11287","display_name":"Exploit","level":2,"score":0.6140999794006348},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.5950000286102295},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.5921000242233276},{"id":"https://openalex.org/C39890363","wikidata":"https://www.wikidata.org/wiki/Q36108","display_name":"Generative grammar","level":2,"score":0.5527999997138977},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.484499990940094},{"id":"https://openalex.org/C90509273","wikidata":"https://www.wikidata.org/wiki/Q11012","display_name":"Robot","level":2,"score":0.4611000120639801},{"id":"https://openalex.org/C166109690","wikidata":"https://www.wikidata.org/wiki/Q4677422","display_name":"Action selection","level":3,"score":0.4546999931335449},{"id":"https://openalex.org/C167966045","wikidata":"https://www.wikidata.org/wiki/Q5532625","display_name":"Generative model","level":3,"score":0.43779999017715454},{"id":"https://openalex.org/C81917197","wikidata":"https://www.wikidata.org/wiki/Q628760","display_name":"Selection (genetic algorithm)","level":2,"score":0.4034000039100647},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.36649999022483826},{"id":"https://openalex.org/C89611455","wikidata":"https://www.wikidata.org/wiki/Q6804646","display_name":"Mechanism (biology)","level":2,"score":0.36570000648498535},{"id":"https://openalex.org/C105339364","wikidata":"https://www.wikidata.org/wiki/Q2297740","display_name":"Software deployment","level":2,"score":0.32989999651908875},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.31679999828338623},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.30239999294281006},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.2800000011920929},{"id":"https://openalex.org/C189950617","wikidata":"https://www.wikidata.org/wiki/Q937228","display_name":"Property (philosophy)","level":2,"score":0.27250000834465027},{"id":"https://openalex.org/C106131492","wikidata":"https://www.wikidata.org/wiki/Q3072260","display_name":"Filter (signal processing)","level":2,"score":0.26930001378059387},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.25760000944137573},{"id":"https://openalex.org/C77618280","wikidata":"https://www.wikidata.org/wiki/Q1155772","display_name":"Scheme (mathematics)","level":2,"score":0.25220000743865967}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.10094","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10094","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.10094","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.10094","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Vision-Language-Action":[0],"(VLA)":[1],"models":[2],"show":[3,188],"strong":[4],"potential":[5],"for":[6,84],"general-purpose":[7],"robotic":[8],"manipulation,":[9],"yet":[10],"their":[11],"closed-loop":[12,193],"reliability":[13,68],"often":[14,33],"degrades":[15],"under":[16],"local":[17],"deployment":[18],"conditions.":[19],"Existing":[20],"evaluations":[21],"typically":[22],"treat":[23],"test":[24],"episodes":[25],"as":[26],"independent":[27],"zero-shot":[28],"trials.":[29],"However,":[30],"real":[31],"robots":[32],"operate":[34],"repeatedly":[35],"in":[36,96,196],"the":[37,89,135,143,149,159],"same":[38],"or":[39],"slowly":[40],"changing":[41],"environments,":[42],"where":[43],"successful":[44,72,93,165],"executions":[45],"provide":[46],"environment-verified":[47],"evidence":[48],"of":[49,142],"reliable":[50],"behavior":[51],"patterns.":[52],"We":[53,75],"study":[54],"this":[55,123],"persistent-deployment":[56],"setting,":[57],"asking":[58],"whether":[59],"a":[60,97],"partially":[61],"competent":[62],"frozen":[63,160],"VLA":[64,161],"can":[65],"improve":[66],"its":[67,71],"by":[69],"reusing":[70],"test-time":[73,81,178],"experience.":[74],"propose":[76],"an":[77,117,139],"online":[78],"success-memory":[79],"guided":[80],"adaptation":[82,179],"framework":[83],"generative":[85,170],"VLAs.":[86],"During":[87],"deployment,":[88],"robot":[90],"stores":[91],"progress-calibrated":[92],"observation-action":[94],"segments":[95],"long-term":[98],"memory.":[99],"At":[100],"inference,":[101],"it":[102],"retrieves":[103],"state-relevant":[104],"action":[105,119,126,145],"chunks,":[106],"filters":[107],"inconsistent":[108],"candidates":[109],"via":[110],"trajectory-level":[111],"consistency,":[112],"and":[113,147,185,192,198],"aggregates":[114],"them":[115],"into":[116,125,138],"elite":[118,136],"prior.":[120],"To":[121],"incorporate":[122],"prior":[124,131,137],"generation,":[127],"we":[128],"introduce":[129],"confidence-adaptive":[130],"guidance,":[132],"which":[133],"injects":[134],"intermediate":[140],"state":[141],"flow-matching":[144],"sampler":[146],"adjusts":[148],"guidance":[150],"strength":[151],"based":[152],"on":[153],"retrieval":[154],"confidence.":[155],"This":[156,172],"design":[157],"allows":[158],"to":[162],"exploit":[163],"environment-specific":[164],"experience":[166],"while":[167],"preserving":[168],"observation-conditioned":[169],"refinement.":[171],"retrieve-then-steer":[173],"mechanism":[174],"enables":[175],"lightweight,":[176],"non-parametric":[177],"without":[180],"requiring":[181],"parameter":[182],"updates.":[183],"Simulation":[184],"real-world":[186],"experiments":[187],"improved":[189],"task":[190],"success":[191],"stability,":[194],"especially":[195],"long-horizon":[197],"multi-stage":[199],"tasks.":[200]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-05-13T00:00:00"}